When Foreign Exchange Becomes Domestic Liquidity
This chapter follows a fixed exchange-rate arrangement through its most important operating channel: how a foreign-exchange transaction changes the central bank balance sheet and the reserve position of commercial banks. The practical goal is to let the reader trace a reserve shock from its first ledger entry to the domestic money-market rate and the capital-flow response, while keeping clear which movements are automatic and which require a policy decision.
From an FX transaction to the central bank balance sheet
A peg is often described as a promise to buy or sell foreign currency at a fixed price. That description is accurate but incomplete. The transaction is also a monetary transaction because the central bank pays for foreign currency with a liability that matters to the banking system: a local-currency reserve balance.
Start with the central bank’s simplified balance sheet. Its foreign reserves are an asset. Its liabilities include currency in circulation, the government’s account, bank reserves, and other items. Commercial banks hold accounts at the central bank. Those accounts are not a decorative record of the peg; they are the settlement balances banks use to make payments among themselves and to meet their own liquidity needs.
Now consider a foreign-currency surplus. A commercial bank may receive foreign currency from exporters or investors, or it may acquire more foreign currency than it wants to keep. Under a fixed exchange-rate arrangement, the bank can sell the foreign currency to the central bank at the committed rate. The central bank adds the foreign currency to its reserves and credits the bank’s local-currency account. Two entries occur together:
- foreign assets on the central bank’s balance sheet increase;
- the commercial banking system’s reserve balance increases.
The local currency has not appeared from nowhere. It has been issued as the counterpart to the foreign asset acquired by the central bank. In a currency-board arrangement, this link is especially explicit: the monetary base is issued only against foreign-currency reserves. A conventional peg may use a broader set of operations, but the same balance-sheet channel still matters whenever the central bank buys foreign currency and pays in local currency.
The reverse transaction is equally important. If banks or their customers need foreign currency, the banks buy it from the central bank and pay with local-currency reserves. The central bank’s foreign assets decline, and bank reserves decline. The exchange rate has not moved, but the domestic liquidity position has tightened.
A purely illustrative ledger makes the direction clear. Suppose a bank sells 100 units of the anchor currency to the central bank at the committed rate. The central bank’s foreign-asset position rises by the local-currency value of those 100 units, while the bank’s reserve account receives the same local-currency amount. If the bank later buys 100 units of foreign currency, the entries reverse. The numbers are only a teaching example; the important fact is the paired movement between the foreign-asset side of the central bank’s balance sheet and the reserve liability owed to banks.
The central bank is therefore not merely defending a price in an FX market. It is changing the quantity of the safest settlement asset available to domestic banks. That is why a peg’s external transactions can influence domestic funding conditions even when the central bank has not announced a change in its policy stance.
The reserve movement also depends on who initiates the transaction and what happens next. A customer’s export receipt may first increase a bank’s foreign-currency holdings. The bank can keep those holdings, sell them to another market participant, or exchange them with the central bank. The peg constrains the price of the last transaction, but it does not guarantee that every foreign-currency flow will be immediate, costless, or routed through the same channel. Settlement timing, market access, and the bank’s own liquidity preference affect when the reserve effect appears.
The accounting relationship becomes economically important because banks do not view reserves as a passive number. They need reserves to settle payments during the day and to manage the risk of being short at the end of the day. If the banking system as a whole has fewer reserves than banks collectively want, some banks must borrow from others or from the central bank. If it has more than banks want, banks will look for a way to lend, invest, or exchange the surplus. The reserve position becomes the starting condition for the domestic money market.
There is a useful distinction between an individual bank’s position and the system’s aggregate position. One bank can be short of reserves while another is long. An interbank loan can redistribute the existing balances without changing the total reserves supplied by the central bank. But if every bank is short because a foreign-currency payment has drained the system, banks cannot solve the aggregate shortage merely by trading among themselves. They can only pass the shortage from one balance sheet to another unless the central bank, a foreign-currency inflow, or some other balance-sheet movement supplies new reserves.
The same logic applies to an aggregate surplus. A bank can lend its excess to another bank, buy an eligible asset, or place funds at a central-bank facility, but those choices do not by themselves make the surplus disappear from the banking system. They change who holds it and what return it earns. The central bank can change the aggregate quantity through an FX transaction, an asset operation, or a liquidity-absorbing instrument. That distinction—redistribution within the banking system versus a change in the system-wide supply—is essential for understanding why an active money market cannot always neutralize a peg-related shock on its own.
A reserve balance also has a time dimension. A bank may have enough funds on average but still face a shortfall at the end of a settlement day. Conversely, a bank may hold a buffer because it cannot be sure that a counterparty will lend tomorrow. The reserve demand reflected in the money market is therefore not a fixed physical requirement. It depends on payment flows, market depth, counterparty confidence, and the cost of being caught short. A fixed exchange rate changes the source and potential size of those flows, but it does not eliminate the banking system’s need to manage them.
The diagram shows a sequence, not a claim that every transaction passes through every participant in this exact order. Its point is that the FX operation and the money-market condition are connected through the central bank balance sheet. The rate response and the capital-flow response can then feed back into the same FX market.
From reserve quantity to rates, flows, and daily noise
When reserves fall, a bank facing payments or settlement needs has fewer immediately available funds. It may borrow reserves from another bank. If many banks are short at once, the price of that borrowing—the short-term money-market rate—tends to rise. When reserves increase, banks are less desperate to borrow and may compete to lend the surplus, putting downward pressure on the rate.
This is a reserve-price channel, not automatically a new policy-rate decision. A central bank can leave its announced policy rate unchanged while an unoffset reserve movement makes the overnight market rate move. The distinction matters for reading a peg. A market rate that rises after a foreign-currency sale is not necessarily evidence that the central bank has chosen a tighter monetary stance. It may be the market’s immediate response to a smaller supply of settlement balances.
The rate movement is also part of the peg’s adjustment mechanism. If the domestic short-term rate rises relative to the anchor-country rate, holding local assets may become more attractive. That can encourage capital inflows. Investors bring foreign currency into the market, and banks may sell that currency to the central bank in exchange for local reserves. The inflow then partly reverses the original reserve shortage and reduces the pressure on the domestic rate.
The opposite sequence follows a reserve surplus. More abundant reserves reduce the incentive to bid for funds. The domestic rate may fall relative to the anchor. If local assets now offer less compensation, capital may flow outward. Banks or investors buy foreign currency, paying with local reserves, and the surplus can contract.
Economists often describe the rate relationship through uncovered interest parity, or UIP. In plain terms, it is the idea that differences between domestic and foreign interest rates are related to the expected movement of the exchange rate. When the exchange rate is fixed, the expected exchange-rate movement is constrained, so the anchor rate becomes an important reference point. In practice, transaction costs, settlement timing, country risk, and market depth create a wedge around that reference. The peg does not make the local rate identical to the anchor rate; it makes deviations consequential for capital flows and the reserve position.
That last distinction protects against a common interpretive error. A higher local rate can be the result of scarcity in the reserve market, a deliberate policy decision, or a risk premium demanded by lenders. All three may appear as a wider domestic-versus-anchor spread, but they imply different responses. If scarcity is temporary, a liquidity operation may be enough. If the rate has been raised to discourage outflows, the central bank is using the price of local funding as part of external defense. If the spread reflects country or counterparty risk, lowering it administratively may not restore confidence and could instead exhaust reserves or encourage hidden leverage.
The rate is therefore both a signal and a transmission channel. It signals how scarce or risky local liquidity appears, while transmitting that condition to banks, investors, and foreign-exchange dealers. The direction of the signal is clear in the stylized model, but its size and timing are empirical questions. A small reserve movement can produce a large rate change in a shallow market; a large movement can produce little response when reserves are abundant or when a standing facility anchors the market. The same balance-sheet entry can consequently have different financial consequences in different operating environments.
The response is not driven by FX flows alone. Reserves can move because of what the IMF framework calls autonomous factors: balance-sheet items that affect bank reserves without being directly controlled by the central bank’s liquidity operation. Currency in circulation is one example. When the public wants more cash, banks exchange reserves for notes, reducing the balances left in the banking system. When cash returns to banks, the direction reverses.
The government’s account at the central bank is another example. Tax payments can move reserves from commercial banks into the government account. Government spending can move them back. The central bank may know the broad pattern, but the timing and size of receipts and payments can create daily forecast errors. Foreign assets and net foreign assets also change with current-account and capital-account flows. Under a peg, those flows can be large relative to a small domestic money market.
The practical implication is not that every movement must be forecast perfectly. It is that the central bank needs to know which movements it can offset, which it can only buffer, and which it must allow the money market to absorb. A forecast that treats the foreign-reserve position as the whole story will miss cash demand and fiscal flows. A forecast that treats the policy rate as the whole story will miss the quantity of settlement balances available to banks. The operating problem is to connect the two views: identify the source of the reserve movement, estimate how banks will react, and decide whether an operation should smooth the path or let the rate carry the signal.
This explains why the reserve position can change even when the central bank has not changed its announced objective and no bank has made a conscious bet against the peg. A public cash withdrawal, a government payment, or a foreign-currency transaction can all alter the quantity of reserves available for settlement. The resulting rate movement may then provoke a capital-flow response. A peg is consequently an operating environment in which the central bank must forecast a moving balance sheet, not simply publish a fixed rate.
There is another limit to the automatic mechanism. A currency-board rule can constrain how local currency is issued, but it cannot guarantee that banks will always find a willing counterparty, that foreign investors will respond instantly to a rate change, or that the domestic market will absorb a shock without stress. The rule determines an important relationship between the monetary base and foreign reserves. It does not remove settlement delays, risk premia, uneven access to funding, or uncertainty about future conditions.
Imagine two economies with the same fixed exchange rate and the same initial reserve shortage. In the first, banks trust one another, the interbank market is deep, and foreign capital responds quickly to a small rate difference. The shortage may produce only a modest, short-lived rate movement. In the second, banks distrust counterparties, the market is shallow, and investors worry about country risk. The same shortage may produce a sharp rate increase that does not bring enough capital inflow. The exchange rate can remain unchanged in both economies, while the second banking system pays a much higher price for the commitment.
The reserve-flow mechanism therefore gives the peg its first line of adjustment, but it does not settle the question of resilience. The next issue is whether the feedback arrives quickly enough and cheaply enough to stabilize the system. Transaction costs, market segmentation, and risk can interrupt the loop; when they do, an exchange-rate commitment may push the pressure into interest rates, bank liquidity, and money-market development instead. The mechanism must be understood before those frictions can be diagnosed.